Biopharmaceutical companies are embracing Industry 4.0 ideas and using digital technologies to revolutionize manufacturing and improve the quality of medicines. Well, at least a few of the larger firms are. Others will take longer, say experts.
Going digital takes resources. Setting up the infrastructure needed to gather data and feed it back to analytical and modeling systems is a complex process that takes time and costs money, says Caterina Minelli, PhD, principal research scientist at the National Physical Laboratory, the national measurement standards laboratory for the United Kingdom.
“Digital manufacturing approaches [encompass technologies such as] artificial intelligence, big data, cloud-based computing, and the Internet of Things (IoT), as well as real-time predictive modeling, robotics, and automation,” she continues. “Some aspects of these elements can be combined to deliver digital connectivity across the entire supply and manufacturing chain, enabling data to be connected and allowing information to flow through the chain and be easily retrieved.
“This connectivity requires a significant data infrastructure underpinned by standardized data ontology, terminology, and instrument interfaces. It also requires the ability to manage big sets of data.”
As a result, biopharma’s move toward digital manufacturing has been led by larger companies that have the capacity to invest. GlaxoSmithKline (GSK), for example, has been working to digitalize its operations for several years. GSK started using digital technology to enhance interactions with patients and consumers. The company is also interested in using digital technology to optimize manufacturing operations.
GSK recently indicated that it is working with Siemens and Atos, two of the world’s leading expert companies in digital transformation and technology, to realize the “vaccine factory of the future.” GSK’s vaccine factory will incorporate a digital twin. (A digital twin is a complete and real-time simulation of an entire manufacturing process. By providing computational models that let scientists test and modify experimental parameters, a digital twin can improve process development more efficiently.)
GSK expects to deploy digital twins more generally. “Traditional process development methods are still being used, but our ambition is to progressively implement digital twins,” says Sandrine Dessoy, digital innovation lead, GSK. “The development of digital twins is an investment not only for R&D, but also for manufacturing and quality control activities.”
GSK’s approach is to work with industrial data management experts to build digital models of candidate processes and then to try to predict how they will perform before they are scaled up and implemented on the factory floor.
“We started with a proof-of-concept project in collaboration with Atos and Siemens to determine how a digital twin for a vaccine production process should be defined, built, and demonstrated,” Dessoy detailed. “The next step was to move from the minimum viable product (MVP) to a robust, scalable future-ready platform—this was the main challenge.”
MVP is the technical term for a system that functions based on the bare minimum of components. According to Dessoy, the “minimum number” for GSK’s digital twin turned out to be quite large.
“Many components, such as fast hybrid models, sensors, process analytical technology, automation, and data streaming and modeling platforms, had to be integrated while a fast data flow, from equipment to model and to automation, was ensured,” she added. “The solution needed to be GMP compliant and flexible enough to be implemented in R&D and manufacturing with a diverse range of equipment.”
Despite these challenges, the digital twin project has been a success. It has given GSK’s vaccine manufacturing team more options, and according to Dessoy, the digital twin approach is being applied to well-established and newly introduced production processes alike.
“We have been running two approaches in parallel,” she relates. “One approach involves the development of a complete digital twin for new vaccines; the other, the implementation of a digital twin for monitoring of lifecycle projects.
“The complete digital twin will control the process and will be transferred to GMP and production alongside the vaccine project. The lifecycle digital twin follows a simpler approach: feed the model with years of data to better monitor the process in real time and detect deviation, but with no regulatory impact. This simpler digital twin is already delivering value in production and increasing capacity.”
So, for companies like GSK that have the requisite resources and time to invest in development, digital manufacturing is becoming a reality. However, for many smaller or less deep-pocketed companies, the adoption of digital technologies is taking place more slowly.
“Different organizations and market subsegments are adopting different elements of the digitalization approach,” explains Minelli, who works with manufacturers from a range of industries on the analysis of uncertainty and variance in digital systems. “There are many who have developed and optimized processes around digital notebooks and wearable technologies like augmented and virtual reality headsets for training and maintenance purposes.
“However, true digital connectivity across the supply chain is still in various stages of development, with a range of new standards and digital tools required to fully realize the benefits available through a complete shift to digital control.”
This view is shared by Richard Vellacott, CEO of BiologIC Technologies, an industrial software and systems developer. He maintains that for mid-sized biopharma firms, the transition from analog production is a daunting prospect: “Biopharma companies are incrementally adopting Industry 3.0 digital manufacturing but will never achieve full-stack Industry 4.0 digital adoption because they have too much interest in defending the status quo.”
Vellacott allows that the situation is different for emerging biotech companies. These companies are building their manufacturing infrastructure from the ground up, so they are more likely to embrace digitalization.
According to Vellacott, the shift to Industry 4.0 will occur in biomanufacturing much like it occurred in other kinds of manufacturing: “Transformative digital manufacturing—because it is so disruptive to incumbents and existing supply chains—will emerge from new companies that have fundamentally reimagined the power of digitalization in what we might call the Kodak moment for therapy.”
Vellacott suggests that the move to full-stack digital manufacturing may be accelerated in light of recent events. “The pandemic has painted a clear picture in the public eye of how slow, expensive, and inflexible our classical biomanufacturing processes are even in response to urgent threats,” he observes. “The industry has to focus on order-of-magnitude improvements. Only full-stack Industry 4.0 digital manufacturing can make order-of-magnitude improvements to therapy cost, the availability of rapid on-demand manufacturing at the point of need, and elastic manufacturing capacity.”
“If we achieve this, we will vastly increase accessibility for patients. The benefits will be transformative. They will include mitigation of pandemics, pervasive adoption of personalized therapies, and far more sophisticated combination therapies.”
Vellacott says that technological solutions are being developed with emerging companies in mind. Indeed, he points out that such solutions are being developed by his own firm.
“BiologIC’s biocomputer is an example of a full-stack digital manufacturing approach,” he declares. “It is based on highly integrated designs enabled by pioneering innovations in additive manufacturing and novel methods of operation, sensing, and data intelligence. Key features include software configurability to enable elastic capacity, digital simulation to allow for novel process development, and machine learning to optimize manufacturing strategies for complex bioproducts.”
Like Vellacott, Minelli is of the opinion that the coronavirus pandemic is bound to accelerate the biopharma industry’s adoption of digital technology. “The pandemic,” she explains, “has necessitated an increased pace of adoption of digitalization solutions across biopharma manufacturing.” She suggests that when companies integrate different systems to solve immediate problems, they set precedents for collaborations of larger scope.
“The pandemic has pushed the pharmaceutical sector workforce to adopt digital methods of communication,” she adds. “This, in turn, has forced manufacturers to ensure that the employees have access to data and analysis tools remotely. Finally, remote access has accelerated the adoption of digital manufacturing techniques. Employees can be productive off the shop floor and spend time generating insights about manufacturing processes.”
Digital technologies will play a greater role in drug production whether they are pioneered by large biopharma companies or emerging biotech firms or both. In any case, regulators will need to prepare for the change.
“Inevitably, technological revolutions happen at a pace far quicker than the response time of regulators and governments, and they will have to respond after the fact,” Vellacott states. “Digital companies cannot slow their pace. Instead, they have to navigate the environment with new strategies that fundamentally assure therapy efficacy and patient safety while leading the public debate based on the need to radically reduce therapy costs and increase their accessibility.
“Forward-thinking governments should invest in Industry 4.0 digital manufacturing technologies if they wish to develop sovereign manufacturing control and supply chain resilience. The regulatory environment is most in flux with the newest cell and gene therapies. I believe that the Food and Drug Administration in the United States and the Medicines and Healthcare products Regulatory Agency in the United Kingdom are responding in the best way they can while always ensuring quality for patients.”